The present invention relates generally to the field of computers, and more particularly to relational database management distributed systems.
Computer memory can be an expensive and scarce resource. As such, memory caching is one of most frequently used methods to improve application performance by cutting down on the loading/unloading of large memory objects, such as tables, indexes, table ingest metadata, table compression metadata etc. The need for keeping such large objects in memory is practical and has been proven to consistently show desired performance results. One of the examples is when a user regularly ingests data into a table while querying the same table. This popular high concurrency scenario fulfills the very practical need for most-up-to-date data analytics.